Each design iteration prerequisites, after a review, a semi-design freeze. I get the point that we might not have the CNC capabilities to create injection molds on scale in Pakistan. No reason behind that is because we don't have a consumer industrial base which is shooting out 100'000s of plastic parts/day.
Now to this idea of procuring 3D-Printers; our import regulations don't support the vast decentralised ABS-parts-farming.
But rather than "Design to Function" we could go for "Design to Process". Which means that you design to your established industrial capacity/capability, without first investing too much in a technology whcih an alien in your local envoirment (aka. 3D-Printing/ABS-cassettes/ .stl files etc ). A repeatable, high-quality workflow like vacuum-forming for example could be explored IMO. Image below is one example. The same result can be replicated using different thickness of pvc-sheets etc.
View attachment 199791
credit:
https://techfoundry.ucdavis.edu/vacuum-forming